Our studies in this project have focused on determining the mechanisms of salivary gland and neural crest formation. We are addressing the following major questions: 1. How do embryonic salivary glands and other tissues generate their large epithelial surface areas during the process of branching morphogenesis? Specifically, how is cleft formation that delineates buds and ducts mediated and regulated, and how do epithelial tissues expand rapidly while remaining constrained by the basement membrane? How can we facilitate bioengineering for organ replacement, particularly of salivary glands, by understanding branching morphogenesis and by developing reconstitution approaches? 2. What are the roles of the regulation of extracellular matrix, signal transduction, selective gene expression, and cell migration in branching morphogenesis and in other major tissue rearrangements such as cranial neural crest development? We are applying a variety of approaches to begin answering these complex questions. These approaches include: laser microdissection; gene expression profiling; RNA interference; whole-embryo, organ, and cell culture; confocal immunofluorescence and brightfield time-lapse microscopy; and a variety of functional inhibition and reconstitution approaches. Salivary glands are formed from the embryonic neural crest and other tissues. Their complex architecture is established by the dramatic process of branching morphogenesis in which a simple bud is transformed rapidly into a complex, branched early organ. This process occurs within days in a mouse embryo. Branching is crucial for the formation of many other organs including lungs, kidneys, and mammary glands. Understanding the mechanisms of branching morphogenesis, and applying this knowledge to control tissue self-assembly and branching, should accelerate tissue engineering approaches to regenerate damaged salivary glands or create an artificial salivary gland and other tissues. We used laser-microdissection with SAGE (serial analysis of gene expression) or whole-genome microarrays to identify novel candidate regulators. We completed data acquisition and general bioinformatics analyses, identifying genes that were co-expressed in specific regions of developing embryonic salivary glands versus those expressed uniquely at just one region, such as the forming epithelial cleft or external epithelial cells of end buds. This microanatomical atlas provides a broad overview of the complex and changing gene expression landscape of gland development. Our primary datasets from this study, including 84 microarrays of laser-microdissected regions, are now available for use by the research community through the GEO database and at http://sgmap.nidcr.nih.gov. As an example of the value of this approach, we identified a novel regulatory change: a substantial reduction in GSK3 expression at forming clefts. We confirmed this finding at the protein level, and global inhibition of GSK3 in developing salivary glands enhanced branching morphogenesis. We expect that other new regulators of salivary development will be discovered using this and other databases. Although these approaches will continue to identify promising candidates for mediators and regulators of gland development, exploration and functional validation of potential roles for these genes has been difficult because of the absence of efficient techniques for genetic manipulation of cells within specific tissues of embryonic organs. We identified new vectors for high-efficiency viral gene transfer and cell-type specificity to developing salivary glands. We compared a number of vectors: adenovirus, lentivirus, and eleven types of adeno-associated viruses (AAV) for their ability to transduce embryonic mouse salivary glands. Two AAV types, AAV2 and bovine AAV (BAAV), were able to target gene expression differentially and selectively to epithelial versus mesenchymal cells, respectively. As a proof-of-principle demonstration, infection of embryonic salivary gland epithelia with a self-complementary (sc) AAV2 virus expressing fibroblast growth factor 7 (Fgf7) supported gland survival and in fact stimulated salivary gland branching morphogenesis. These findings represent the first successful selective gene targeting to epithelial versus mesenchymal cells in an organ undergoing branching morphogenesis. Another set of studies has recently identified anosmin as a novel regulatory protein necessary for FGF, BMP, and WNT signaling, cranial neural crest formation, and craniofacial morphogenesis. These studies have focused on cranial neural crest cells, which comprise a stem cell-like population that generates craniofacial bones, teeth, salivary glands, and surrounding connective tissues. During embryonic development, the neural crest is formed at the boundary of the epidermal ectoderm and the neural ectoderm. Although many growth and transcription factors are known to regulate neural crest formation and dispersal of neural crest cells to target tissues, extracellular matrix molecules were generally thought to play only downstream roles, e.g., providing fibronectin-rich or basement membrane pathways for neural crest cell emigration. In a high-risk project, we searched for new extracellular matrix regulators of neural crest formation. We hoped to find a matrix molecule that might play a regulatory role in modulating the growth factor and transcriptional pathways essential for neural crest formation. Using microarray screening and other approaches, we discovered that anosmin shows strong differential expression in the early chick neural crest, the most popular model system for analyzing neural crest function. RNA interference, overexpression, and protein microinjection studies established that anosmin plays essential roles in neural crest function, with striking effects on the generation of neural crest cells and craniofacial morphogenesis. Intriguingly, anosmin functions at the intersection of major FGF, BMP, and WNT signaling pathways involved in neural crest formation. Anosmin could directly modulate the activities of FGF8, BMP5, and WNT3a, as demonstrated by direct reporter assays using purified molecules and in vivo signaling. Importantly, knockdown of anosmin expression using morpholino oligonucleotides strongly inhibited neural crest formation, and development could be rescued successfully using an appropriate balance of elevated levels of FGF8 and BMP5; the latter growth factor was also identified as a new regulator of neural crest formation in addition to BMP4. Experimentally altering levels of anosmin induced subsequent craniofacial defects and altered maxillary and mandibular processes in developing embryos. These findings provide direct evidence that a single extracellular matrix protein can play crucial roles in morphogenesis by modulating the balance of multiple growth factor activity-receptor functions.